A Sociolinguistic Analysis of Dzongkha: Variation in Final Nasals and Rhotics
Wangchuk Rinzin
A thesis submitted for the degree of Doctor of Philosophy Department of Language and Linguistics University of Essex (December 2018)
Acknowledgements Page | i
Acknowledgements The writing of this sociolinguistic thesis would have never been completed without the untiring guidance of my supervisor, committee members, backing from my university and parent office, help from my friends, and timeless support from my family.
At this stage, I would like to extend a very heartfelt and deepest gratitude to my supervisor, Professor Peter L. Patrick, for his excellent guidance, patience, caring, suitable comments, distinctive advice, encouragement and genuine support from the beginning to the final systematic version. He enabled me to develop an understanding of this very challenging project.
I am also incredibly grateful to my committee chair Professor Enam Al-Wer, who is well- versed in the sociolinguistic field and believed in my research and work from the beginning to the end. Similarly, I am also very much indebted to my advisor Dr. Vineeta Chand for her invaluable teaching on this particular subject with advice and suggestions during the past forty- four months.
I am very thankful to Dr. Uri Horesh, Dr. Khairiah Al-Qahtani and Dr. Kosin Panyaatasin for helping me in Rbrul and answering a number of my queries related to statistical issues.
My appreciation also goes to Miss. Siham Rouabah for reading the draft of my thesis, invaluable encouragement, helping my family during my absence and continuous support during the busiest times.
From the department of langauge and linguistics, I would like to thank Dr. Wyn Johnson for enlightening me on theoretical and descriptive phonology and Dr. Desmond Thomas for giving me opportunities to practise with PhD thesis writing part I, II, and III, and other helpful and respectful colleagues for their relentless service throughout my study.
Acknowledgements Page | ii
I would also like to express my sincere gratitude to my government (The Royal
Government of Bhutan) and my parent office (Dzongkha Development Commission), for helping me to obtain study leave from the office for the period of four years.
My gratitude also goes to the UK-Bhutan Society for their vital generosity in supporting this survey with the amount of £500 for supplementary expenditure during data collection.
I have also not forgotten to thank all my valuable informants who kindly spent their precious time when participating in this study. Notably, they trusted me and agreed to be recorded and present my thesis with their real names.
I would like to express my great appreciation for my family who have shared the bliss and stress incurred during the writing of this thesis. I am also exceptionally grateful to my son,
H.E. Truelku Ugyen Drodrul Thinley Kunchap, for performing Serkem and Soelkha (offering of celestial drinks) to immediate and ultimate deities for the success of this PhD thesis.
Finally, but by no means least, I offer warmest regards and finest wishes from the bottom of heart, Tashi Delek ‘auspicious-goodness’, to all of those who have offered a hand to me in any circumstances through the completion of the task. Thus, I dedicate this thesis to all of you! Abstract Page | iii
Abstract
This first quantitative sociolinguistic analysis of Dzongkha (Bhutan’s official language), as spoken by residents of the capital Thimphu, investigates variation and change in two salient and traditional linguistic features: syllable-final nasals (N) and postvocalic rhotics (R). Thimphu is Bhutan’s central location for education, jobs, commerce and social network ties. Both (N) and
(R) show variable deletion, as correlated with internal (phonetic environment, tone, grammatical category, phrase position) and external (style, sex, age, region, education) explanatory factors.
Data came from thirty-six participants originating in three regional communities (Eastern native Tshangla speakers, Western native Dzongkha speakers and Southern native Lhotshampa speakers). All were Bhutanese nationals now living in Thimphu, divided amongst school children at seven schools, their teachers and their parents. 3,636 nasal tokens and 2,196 rhotic tokens were analysed using Rbrul to perform multiple logistic regression. The findings demonstrate variation (and suggest change in progress) for both (N) and (R).
Low and mid vowels, prepositions and adjectives, preceding and following sonorants, low-toned syllables, and non-initial position favour deletion of (N). Lhotshampa and Dzongkha speakers delete nasals more; speakers with secondary education preferred the traditional form.
For postvocalic (R), among linguistic factors, low-toned syllables, certain grammatical categories, preceding front vowels, non-final positions and following obstruents promote deletion. Speakers with Western Dzongkha backgrounds favoured deletion, as did older adults generally. For both (R) and (N), principal results showed final consonants retained in formal reading tasks, but style could not be included in multiple regression analysis.
Variation in Dzongkha reflects external developments and socio-economic changes across Thimphu and the country in recent decades. Qualitative analysis of linguistic attitudes, ideology and identity also contribute towards explaining variation and potential change in the use of these features. A grammatical sketch and history of Dzongkha are provided.
Keywords:
Bhutan, Dragon Kingdom, Druk, Dzongkha, Official Language, postvocalic rhotics, primary schools, Rbrul, sociolinguistics, syllable-final nasals, variation and change. Table of Contents Page | iv
Table of Contents Acknowledgements ...... i
Abstract………………………………………………………………………………………..iii
Table of Contents ...... iv
List of Tables ...... xi
List of Figures ...... xiv
List of Maps …………………………………………………………………………………..xv
List of Abbreviations ...... xvi
Note on Transcription ...... xix
Chapter 1 Introduction ...... 1
1.1 Overview ...... 1
1.2 Motivation and Aim of the Study ...... 3
1.2.1 The Context of Language Planning and Policy in Asia ...... 3 1.3 Significance of the Study ...... 7
1.4 Structure of the Thesis ...... 7
1.5 Research Questions and Hypothesises ...... 8
1.6 Diglossic Situation ...... 10
1.7 Geographic and Demographic Information ...... 12
1.8 Maps and Dialect Classifications in Bhutan ...... 16
1.9 Multilingualism in Thimphu City ...... 18
1.10 Linguistic History of Dzongkha ...... 19
1.11 Linguistic Family of Dzongkha ...... 21
1.11.1 Tibeto-Burman or Sino-Tibetan Family and the Position of Dzongkha ...... 21 1.11.2 Old, Middle and Modern Dzongkha ...... 22 1.12 Language Policy in Bhutan ...... 24
1.12.1 Brief Account of Language Planning and the Dzongkha Development Commission…………………………………………………………………………………24 1.12.2 Lingua Franca, Official Status and Decrees of Dragon Kings ...... 26 1.12.3 Dzongkha Curriculum in Schools ...... 27 Table of Contents Page | v
1.12.4 The Current Standing of Dzongkha ...... 29 Chapter 2 Literature Review ...... 33
2.1 Overview ...... 33
2.2 Brief Account on Language Ideology and Behaviour in Bhutan ...... 33
2.3 Speech Community ...... 34
2.3.1 The Concept of Speech Community ...... 34 2.3.2 Preliminary Introduction on Ethnicity and Language ...... 37 2.4 Distinctive Features in Dzongkha Speech Among Different Social Groups ... 38
2.5 A Brief Phonetic Description of Dzongkha Consonants and Vowels ...... 44
2.5.1 Dzongkha Consonant Inventory ...... 44 2.5.2 Dzongkha Vowel Inventory ...... 49 2.5.3 Dzongkha Inherent Vowel ...... 51 2.5.4 Tones and Tone Contour System in Dzongkha Words ...... 52 Chapter 3 Research Methodology ...... 55
3.1 Overview ...... 55
3.2 Choice of Research Site: Geographical and Social Distribution ...... 55
3.2.1 Thimphu Dzongkhag (District) ...... 55 3.2.2 Thimphu City Schools ...... 57 3.2.2.1 Zilukha Lower Secondary School ...... 57 3.2.2.2 Jigme Losel Primary School ...... 57 3.2.2.3 Zilnon Namgyaling Primary School ...... 58 3.2.3 Thimphu Regional Schools ...... 58 3.2.3.1 Khasadrapchu Middle Secondary School ...... 58 3.2.3.2 Yangchen Gatshel Lower Secondary School ...... 59 3.2.3.3 Hongtsho Primary School ...... 59 3.2.3.4 Kuzhugchen Middle Secondary School ...... 60 3.3 Community of Practice ...... 60
3.3.1 Social Relations within the Community ...... 60 3.4 Sampling Design ...... 62
3.4.1 The Sampling Method in the Present Study ...... 62 3.4.2 Sampling Stratification ...... 65 3.4.2.1 Age of Participants ...... 66 3.4.2.2 Gender of Participants ...... 68 Table of Contents Page | vi
3.4.2.3 Education Level or Class of Participants ...... 72 3.4.2.3.1 Correlation of Educational Level and Occupational Hierarchy……………...74 3.4.2.4 Ethnic Identity of Participants ...... 76 3.5 The Participant Sample Design of the Study ...... 79 3.6 Data Collection and Analysis ...... 81
3.6.1 The Recording Instrument ...... 81 3.6.2 The Sociolinguistic Interview ...... 82 3.6.2.1 Casual Speech and Sociolinguistic Interviews ...... 85 3.6.2.2 Picture Task ...... 86 3.6.2.3 Minimal Pairs ...... 87 3.6.2.4 Storytelling ...... 87 3.6.2.5 Reading Short Passages ...... 88 3.6.2.6 Language Ideology and Attitudes Towards Dzongkha ...... 89 3.7 The Brief Pilot Interview ...... 89
3.8 Preparation of the data for Analysis ...... 90
3.8.1 Data transcription ...... 90 3.8.2 Coding Method for Variants and Variables ...... 90 3.8.3 Extracting Target Tokens for (N) and (R) Variables ...... 91 3.9 Description of Statistical Software ...... 92
3.10 The Linguistic Variables ...... 93
Chapter 4 The Realisation of Final Nasal (N) in Dzongkha ...... 95
4.1 Overview ...... 95
4.2 A Brief Description of Nasal Codas in Dzongkha ...... 95
4.2.1 Final Nasals in Dzongkha ...... 95 4.2.2 Place and Manner of Articulation for Nasal Endings ...... 96 4.3 A Brief Historical Description of Final Nasals in Tibeto-Burman Languages 97
4.3.1 Some Studies on Nasal Endings in Mandarin ...... 97 4.3.1.1 The Logic of the Three-Step Model in Denasalisation ...... 99 4.3.1.2 The 4-Stage Model in [De]nasalisation ...... 101 4.3.1.3 The Most Appropriate Model for the [De]nasalisation Process in Dzongkha………………………………………………………………………………..102 4.4 An Estimation of the Frequency of [De]nasalisation in Dzongkha ...... 103
4.4.1 Frequency of [De]nasalisation After Five Vowels ...... 103 4.4.2 Linguistic Constraints on [De]nasalization ...... 110 Table of Contents Page | vii
4.5 The Coding Protocol for Dzongkha Nasal Codas ...... 110
4.5.1 The Linguistic and Social Factor Groups ...... 111 4.5.1.1 The Preceding and the Following Segments (Consonants) ...... 111 4.5.1.2 Immediately preceding-1 Segment (Vowels) ...... 115 4.5.1.3 Lexical Tone ...... 116 4.5.1.4 Phrase Position ...... 117 4.5.1.5 Grammatical Category ...... 117 4.5.1.6 Social Explanatory Variables ...... 117 4.6 Multiple Logistic Regression Analysis: Rbrul ...... 119
4.6.1 Rbrul Modeling: The Four Models ...... 119 4.6.2 Evaluating the Four Models ...... 121 4.6.3 Findings and Interpretation ...... 121 4.6.3.1. Dzongkha Nasal Endings and Linguistic Variables ...... 127 4.6.3.1.1 Nasal Endings and Immediately Preceding-1 (Vowel) ...... 127 4.6.3.1.2 Nasal Endings and Grammatical Category ...... 130 4.6.3.1.3 Nasal Endings and Following Segment ...... 134 4.6.3.1.4 Nasal Endings and Lexical Tone (High and Low) ...... 136 4.6.3.1.5 Nasal Endings and Phrase Position ...... 137 4.6.3.1.6 Nasal Endings and Preceding-2 (Consonant) ...... 138 4.6.3.1.7 The role of sonority ...... 140 4.6.3.2 Nasal Endings and Social Variables ...... 141 4.6.3.2.1 Nasal Endings and Mother Tongue ...... 141 4.6.3.2.2 Nasal Endings and Education Level (Class of Participants)...... 144 4.6.4 Cross-Tabulation of Age, Gender and Nasal Deletion ...... 147 4.6.5 Style: Nasals in Text Reading and Casual Speech ...... 151 4.7 Summary of the Results ...... 156
Chapter 5 The Realisation of Postvocalic (R) in Dzongkha ...... 158
5.1 Overview ...... 158
5.2 A Brief Description of Rhoticity in Dzongkha ...... 158
5.2.1 Final Rhotics in Dzongkha ...... 158 5.2.2 (R) Position in Speech Production ...... 160 5.3 Motivations and Significance of the (R) Study ...... 160
5.4 A Brief Description of Postvocalic (R) in Other Related Literatures ...... 161
5.4.1 Studies on Final Rhotics in English Varieties ...... 161 Table of Contents Page | viii
5.5 Constraints on the (R) Variable ...... 164
5.5.1 Linguistic Constraints on (R) ...... 164 5.5.1.1 (R) in Trill and Flap Alternation ...... 164 5.5.2 Social Constraints on (R) ...... 165 5.6 The Coding Practice for Dzongkha Rhotics ...... 165
5.7 Multiple Logistic Regression Analysis: Rbrul ...... 166
5.7.1 Rbrul Modelling: The Four Models ...... 166 5.7.2 Evaluating the Five Models ...... 168 5.7.3 Findings and Interpretation ...... 169 5.7.3.1 Dzongkha Postvocalic (r) and Linguistic Variables ...... 172 5.7.3.1.1 Postvocalic (r) and Lexical Tone ...... 172 5.7.3.1.2 Postvocalic (r) and Grammatical Category ...... 174 5.7.3.1.3 Postvocalic (r) and Preceding-1 (Vowel)…………………………………...175 5.7.3.1.4 Postvocalic (r) and Phrase Position ...... 177 5.7.3.1.5 Postvocalic (r) and Following Segment (Consonant) ...... 178 5.7.3.2 Postvocalic (r) and Social Variables ...... 181 5.7.3.2.1 Postvocalic (r) and Mother Tongue ...... 181 5.7.3.2.2 Postvocalic (r) and Speakers’ Age ...... 185 5.7.4 Cross-Tabulation of Gender and Rhotic Deletion ...... 191 5.7.5 Postvocalic (r) and Speaker Style ...... 193 5.8 Summary of the Results ...... 197
Chapter 6 Conclusion ...... 199
6.1 The Realisation of Nasal Final (N) in Dzongkha ...... 200
6.1.1 Nasal Final (N) Correlated with Preceding-1 (Vowel) ...... 201 6.1.2 Nasal Final (N) Correlated with Grammatical Category ...... 202 6.1.3 Nasal Final (N) Correlated with Following Segment (Consonant) ...... 202 6.1.4 Nasal Final (N) Correlated with Lexical Tone ...... 202 6.1.5 Nasal Final (N) Correlated with Phrase Position ...... 203 6.1.6 Nasal Final (N) Correlated with Preceding-2 (Consonant) ...... 203 6.1.7 Nasal Final (N) Correlated with Mother Tongue ...... 204 6.1.8 Nasal Final (N) Correlated with Education Level ...... 206 6.1.9 General Findings for (N) variable ...... 207 6.2 Realization of Postvocalic (R) in Dzongkha ...... 208
6.2.1 Postvocalic (R) Correlated with Lexical Tone ...... 209 Table of Contents Page | ix
6.2.2 Postvocalic (R) Correlated with Grammatical Category ...... 210 6.2.3 Postvocalic (R) Correlated with Preceding-1 (Vowel) ...... 210 6.2.4 Postvocalic (R) Correlated with Phrase Position ...... 210 6.2.5 Postvocalic (R) Correlated with Following Segment (Consonant) ...... 211 6.2.6 Postvocalic (R) Correlated with Speaker Style ...... 211 6.2.7 Postvocalic (R) Correlated with Mother Tongue ...... 212 6.2.8 Postvocalic (R) Correlated with Speakers’ Age, Gender and Origin ...... 213 6.2.9 General Findings for (R) Variable ...... 214 6.3 Is Thimphu a Single Speech Community? ...... 215
6.4 Discussion of Ethnic Variation ...... 216
6.5 The Social Implication of Syllable-final Nasals and Rhotics for the Education System ...... 219
6.6 Encountered Difficulties with This Study ...... 220
6.7 Suggestions for Further Research ...... 221
Bibliography ...... 223
Appendix 1: The classification of Bhutanese languages, their location and number of speakers (RAOnline 200-2016) ...... 243
Appendix 2: Early Historical Records of Dzongkha and Its Speakers ...... 246
Appendix 3: The Royal Edict issued by the Third King ...... 248
Appendix 4: The Royal Edict issued by the Fourth King ...... 249
Appendix 5: Dzongkha Language Family ...... 250
Appendix 6: Sociolinguistic Interview Modules for PhD research project ...... 256
Appendix 7: Picture Tasks for (N) and (R) ...... 261
Appendix 8: Minimal Pairs for (N) and (R) ...... 265
Appendix 9: Reading Passages for (N) and (R) ...... 268
Appendix 10: Questionnaire for Language Ideology and Attitude ...... 271
Appendix 11: Coding Sheet for (N and R) ...... 277
Appendix 12: Cross-tabulation for (N) by Phonetic Context ...... 296
Table of Contents Page | x
Appendix 13: Cross-tabulation for (R) by Phonetic Context ...... 300
Appendix 14: Cross-tabulation of gender and educational level against rhotics ...... 304 List of Tables Page | xi
List of Tables Table 1.1: List of languages spoken in Bhutan (Gutman & Avanzati 2014: 1) ...... 13 Table 1.2: The example of different forms in old and middle Dzongkha to the modern Dzongkha (Dorji 2011: 21-37) ...... 24 Table 2.1: The sample of thirty Dzongkha consonants ...... 45 Table 2.2: The phonetic inventory of Dzongkha consonants (Downs 2011: 12) ...... 46 Table 2.3: Phonetic inventory of Dzongkha consonants (Hansen 2012: 5) ...... 47 Table 2.4: Spoken Dzongkha represented by initial consonants (Sherpa et al. 2008: 2) ...... 48 Table 2.5: The Dzongkha initial consonants (Mazaudon & Michailovsky 1988: 117) ...... 48 Table 2.6: The examples of contour opposition in low-toned register (Mazaudon & Michailovsky 1988:119)………………………………………………………………………………………...54 Table 2.7: The examples of contour opposition in high-toned register (Mazaudon & Michailovsky 1988: 119) ...... 54 Table 3.1: Stratified sample displaying the age, gender, education level (social categories), geo- ethnic groups, roles and number of speakers ...... 81 Table 3.2: Orthographical and phonological transcription for data transcription ...... 90 Table 4.1: The frequency of nasalisation and denasalisation after 5 vowels uttered by 18 speakers in the dataset……………………………………………………………………………………107 Table 4.2: The table of implicational scale in the frequency of [De]nasalisation ...... 108 Table 4.3: Percentage distribution of nasal endings between its four linguistic variants ...... 110 Table 4.4: The list of all occurring preceding and following segments...... 111 Table 4.5: Model III: Final Rbrul results of the correlation between the use of nasal codas (N) and the significant independent variables among the community of Dzongkha speakers with various factor groups...... 122 Table 4.6: Cross-tabulation of grammatical category against the (N) variables ...... 125 Table 4.7: Cross-tabulation of phrase position against the (N) variables ...... 126 Table 4.8: Cross-tabulation of lexical tone against the (N) variables ...... 126 Table 4.9: Rbrul results of the correlation between the use of n-deleted variant (Ø) and grammatical category in the speech of Thimphu residents (R2 =0.367; p<0.0001)...... 131 Table 4.10: The super-categories (red in colour) and sub-categories (black in colour) of grammatical categories employed in the first stages...... 132 Table 4.11: Cross-tabulation of preceding-1 (vowel) and grammatical category in the use of n- deletion………………………………………………………………………………………….133 Table 4.12: Cross-tabulation of following segment (consonant) and grammatical category in the use of n-deletion...... 133 Table 4.13: Cross-tabulation of lexical tone (High and Low) and grammatical category in the use of n-deletion…………………………………………………………………………………….134 Table 4.14: Rbrul results of the correlation between nasal deleted variant (Ø) and following segment in the speech of Thimphu residents (R2 =0.367; p<0.0001)...... 135 Table 4.15: Cross-tabulation of two following segments (S & O) against the (N) variables ..... 135 List of Tables Page | xii
Table 4.16: Rbrul results of the correlation between nasal deleted variant (Ø) and lexical tone in the speech of Thimphu residents (R2 =0.367; p<0.0001)...... 136 Table 4.17: Rbrul results of the correlation between nasal deleted variant (Ø) and phrase position in the speech of Thimphu residents (R2 =0.367; p<0.0001)...... 137 Table 4.18: Rbrul results of the correlation between nasal deleted variant (Ø) and preceding-2 segments in the speech of Thimphu residents (R2 =0.367; p<0.0001)...... 138 Table 4.19: Rbrul results of the correlation between nasal deleted variant (Ø) and three major languages spoken across Thimphu residents (R2 =0.367; p<0.0001)...... 142 Table 4.20: Cross-tabulation of three geo-ethnic groups and frequency of realisation in use of four nasal (N) variants (Ø, m, n & ŋ) ...... 143 Table 4.21: Rbrul results of the correlation between nasal deleted variant (Ø) and educational level or social status in the speech of Thimphu residents (R2 =0.367; p<0.05)...... 145 Table 4.22: Cross-tabulation of age and gender in the use of nasal deletion (Ø) ...... 147 Table 4.23: Selected Rbrul results of the correlation between nasal deleted variant (Ø) and data eliciting styles (story, interview and reading) in the speech of Thimphu residents...... 153 Table 4.24: Cross-tabulation between age and style by (N) variants for all speakers ...... 153 Table 5.1: Percentage distribution of (R) between its three linguistic variants ...... 164 Table 5.2: Model V: Final Rbrul results of the correlation between the deletion of word-final (r) and the significant independent variables among the community of Dzongkha speakers with various factor groups...... 170 Table 5.3: Rbrul results of the correlation between the use of r-deleted variant (Ø) and lexical tone in the speech of Thimphu residents (R2 =0.357; p<0.0001)...... 173 Table 5.4: Rbrul results of the correlation between the use of n-deleted variant (Ø) and grammatical category in the speech of Thimphu residents (R2 =0.357; p<0.0001)...... 174 Table 5.5: Rbrul results of the correlation between the use of r-deleted variant (Ø) and preceding-1 (vowel) in the speech of Thimphu residents (R2 =0.357; p<0.0001)...... 175 Table 5.6: Rbrul results of the correlation between r-deleted variant (Ø) and phrase position in the speech of Thimphu residents (R2 =0.357; p p<0.0001)...... 177 Table 5.7: Rbrul results of the correlation between the use of r-deleted variant (Ø) and following segment (consonant) in the speech of Thimphu residents (R2 =0.357; p<0.05)...... 178 Table 5.8: Cross-tabulation of following segments against the (R) variable ...... 178 Table 5.9: Rbrul results of the correlation between the use of r-deleted variant (Ø) and mother tongue across Thimphu residents (R2 =0.357; p<0.0001)...... 182 Table 5.10: Cross-tabulation of the realisation [r] variants by mother tongue or different ethnic groups…………………………………………………………………………………………...183 Table 5.11: Cross-tabulation of the deletion of [r] by mother tongue and age group ...... 184 Table 5.12: Rbrul results of the correlation between the use of r-deleted variant (Ø) and speakers’age across Thimphu residents (R2 =0.357; p<0.05)...... 187 Table 5.13: Cross-tabulation of the use of [r] variants by age and social groups ...... 188 Table 5.14: Cross-tabulation of the use of [r] variants by gender and age groups ...... 189 Table 5.15: Cross-tabulation of gender against rhotic variants by the whole community speakers…………………………………………………………………………………………192 List of Tables Page | xiii
Table 5.16: The selected Rbrul results of the correlation between the use of r-deleted variant (Ø) and style across all Thimphu residents...... 193 Table 5.17: Cross-tabulation between age and styles by (R) variants for all speakers ...... 195 List of Figures Page | xiv
List of Figures Figure 2.1: The frequency of unique Dzongkha-features used among Bhutanese Dzongkha- speakers…………………………………………………………………………………………..40 Figure 2.2: The Dzongkha vowels used for the present study, depicted by author (16/07/2019). 51 Figure 4.1: The three-stage model of denasalisation in Spanish (Poplack 1979), cited in Barale (1982: 10)……………………………………………………………………………………….100 Figure 4.2: The three-step process of denasalisation in Mandarin (Barale 1982: 106) ...... 101 Figure 4.3: The 4-step process of nasalisation (Barale 1982: 147) ...... 102 Figure 4.4: Effects of sonorous environments on nasal deletion, based on cross-tabulation of preceding-2 and following segment ...... 139 Figure 4.5: Sonority hierarchy (Tsunoda 2008: 151) ...... 140 Figure 5.1: Effects of obstruent environments on rhotic deletion, based on cross-tabulation of (R) variable and following segment...... 179 Figure 5.2: Cross-tabulation of the use of r-deletion by gender, origin and age group ...... 190 Figure 5.3: Cross-tabulation of the use of r-deletion by gender, origin and age group ...... 191 List of Maps Page | xv
List of Maps Map 1.1: Map of the major languages of Bhutan with the regions of their speakers, from the Ethnologue: Languages of the world (Simons & Fennig 2018) ...... 17 Map 1.2: Map of the 20 districts of Bhutan (RAonline 2000-2016) ...... 18 Map 1.3: Regions in Bhutan whereDzongkha is natively spoken are highlighted with light- yellow with colour and light white with black and white print, from Wikipedia, the free encyclopedia (2001-2016) ...... 22 List of Abbreviations Page | xvi
List of Abbreviations The following table gives the list of standard abbreviations used for glossing in this thesis. 1 first person
3 third person
ABL Ablative
ADJ Adjective
AUX Auxiliary
CAUS Causative
DEM Demonstrative
DET Determiner
FUT Future
GEN Genitive
IND Indicative
INDF Indefinite
INS Instrumental
NEG Negation, negative
PL Plural
POSS Possessive
PRS Present
PST Past
PTCP Participle
Q Question particle/marker
SG Singular
List of Abbreviations Page | xvii
List of abbreviations for proper nouns and other well-known names used in this document. A: Adult ABS: Australian Burueau of Statistics ADM: Administrative Officer ALDO: Assistant Language Development Officer BFA: Bhutan Film Association C: Child CBS: Centre for Bhutan Studies CLDO: Chief Language Development Officer D: Degree D: Denasalized D: Derhoticised D: Dzongkha DCLDO: Deputy Chief Language Development Officer DCRC: Department of Civil Registration and Census DCRD: Department of Curriculum and Research Development DDA: Dzongkha Development Authority DDC: Dzongkha Development Commission DEO: District Education Officer DIT: Department of Information Technology DSAC: Dzongkha Special Advisory Committee E: East E: Elementary E: Executive EDz: Eastern Dzongkha F: Female GIP: Glassgow Indicators Project GNHC: Gross National Happiness Commission GNHR: Gross National Happiness Recearch H: High ID: Identity L: Lhotshampa L: Low LDO: Language Development Officer M: Male M: Minimal pairs MoE: Ministry of Education MoHCA: Ministry of Home and Culture Affair MoIC: Ministry of Information and Communications MPAB: The Motion Picture Association of Bhutan N: Nasalized NCWC: National Commission for Women and Children NISRA: Northern Ireland Statistics and Research Agency NLB: National Library of Bhutan NRS: National Records of Scotland NSB: National Statistics Bureau O: Old OCC: Office of the Census Commissioner P: Parent P: Picture task
List of Abbreviations Page | xviii
P: Primary P: Professional PMKBUN: Permanent Mission of the Kingdom of Bhutan to the United Nation R: Reading passage R: Rhoticised RCSC: Royal Civil Service Commission REC: Royal Education Centre RGoB: Royal Government of Bhutan PRC: Pew Research Centre S: Secondary S: South S: Story S: Student S: Support SDz: Southern Dzongkha SLDO: Senior Language Development Officer SP: Speaker T: Teacher T: Tshangla TCB: Tourism Council of Bhutan TCC: Thimphu City Corporation TEO: Thromde Education Officer UNGEGN: United Nations Group of Experts on Geographical Names US: Underlying spelling W: West WDz: Western Dzongkha Y: Young
List of Phonetic Transcription Page | xix
Note on Transcription IPA transcription will be used throughout this thesis, and in places will be related to the orthographic symbols used in writing Dzongkha. The inventory of consonants and of vowels is provided in §2.5, and the coding of segments used for Rbrul analysis is discussed in §4.5 and given in full in coding sheets (Appendix 11).
Obstruents: a – [Ɂ] b – [ba] B –[ʈʰ] c – [tʃ] C – [tʃʰ] d – [d] D – [ɖ] g – [g] h – [h] k – [k] K – [kʰ] p – p] P – [ɸ] Q – [q] s – [s] S – [ʃ] t – [t] T – [θ] z – [z] Z – [dʒ] 2 – [dz] 3 – [ʒ] 4 – [ɦ] 6 – [ts] 7 – [ʈ] 8 – [tsʰ]
Sonorants: G – [ŋ] H – [ɬ] l – [l] L – [l] m – [m] M – [m] n – [n] N – [ɳ] r – [r] R – [ɴ] w – [w]
List of Phonetic Transcription Page | xx
W – [w] y – [j] Y – [j] 9 – [ɲ]
Vowels/sonorants i – [i] u – [u] e – [e] o – [o] U – [ʉ] O – [ö] V – [ʌ]/[ɑ]
Other Symbols:
˦ High tone ˨ Low tone ʰ Aspirated ẽ Nasalized Ø/0 Null
Chapter One Page | 1
Chapter 1 Introduction
1.1 Overview This present study investigates sociolinguistic variation and potential change in
Dzongkha, specifically the realisation of nasality and rhoticity among residents of Thimphu, the capital city of Bhutan, focusing on the city (urban) primary schools and regional (rural) primary schools. Nasal deletion is a salient sociolinguistic feature often debated in social media (see 1.3 below), and explicitly discussed by teachers in Dzongkha classes; and it allows a point of comparison with other Sino-Tibetan languages where similar historical procedures occur.
Rhotics are traditionally valued in the written and literary language but are undergoing change in the capital city, the focus of chapter 5. Since both of these features appear to be in the mid- course of change (43%), and deletion of nasal, (see 4.4.2 below); 61% deletion of rhotics,
5.5.1.1), and since the study examines a range of complex social factors which might constrain or promote variation, the thesis adopts a variationist approach.
Dzongkha is the official language of the Dragon Kingdom of Bhutan as cited in the constitution (2008), and also a lingua franca in this multilingual nation. Dzongkha is taught in schools and used in courts and military services (Dorji 1990, DDC 1999, Tshewang 2013). It is a member of the Bodish group of the Western Tibeto-Burman (TB) family, which is a subgroup of the Sino-Tibetan (ST) family (Benedict 1972, Bradley 1997, Matisoff 2003). To date, only a handful of general linguistic studies have been carried out in the country: for example, Byrne
(1909), Mazaudon & Michailovsky (1988), van Driem (1992), Thinley (2002), Sherpa et al.
(2008), Gelles (2010), Hasen (2010), Downs (2011) and Watters (2018).
Hitherto, no sociolinguistic studies have been carried out on Dzongkha or other languages or dialects in the country. This study investigates the use of two linguistic variables: the final nasals ([n], [m], [ŋ] and the null or deleted variant [Ø]) and final rhotics ([r], [ɾ], [ɹ], [ʀ] and the null or deleted variant [Ø]). It considers eight linguistic independent variables (three
Chapter One Page | 2 preceding segments, the following segment, lexical tone, following tone, phrase position and grammatical category) and seven independent social variables (speaker style, role, speaker sex, age, education level or class of participants, origin of speakers and mother tongue). Notably, speakers from different geographical regions of Bhutan typically have distinct ethnic origins, and different relations to Dzongkha, with Western residents historically being privileged as native
Dzongkha speakers. The research also briefly describes language ideology and attitudes towards
Dzongkha, though no detailed survey was carried out.
Given the crucial importance of educational issues, including the use of Dzongkha (see
1.12 below), the sample was organised around the three categories of students, parents and teachers, though it is recognised that this prevents a full sampling of occupational and age ranges. The data were elicited through standard sociolinguistic devices: sociolinguistic interview modules, reading passages, minimal pairs reading tasks, picture tasks and language ideological questionnaires, in seven different primary schools located in the city and region under the
Thimphu district, Bhutan. This study uses a sample of thirty-six speakers: twelve primary school students, twelve primary school teachers and twelve parents with different levels of education, in and around the capital. Being strategically located in the western central part of Bhutan, the political and economic centre of the country, and the location of the central government,
Thimphu gathers people from all walks of life in Bhutan.
Previous works on Dzongkha are generally lacking in some aspects of concrete, scientific and sociolinguistic investigation. This research employs quantitative sociolinguistic methods and qualitative investigation in efforts to empirically study the linguistic processes at work. The research questions are discussed in detail below (§1.5), after the context of Dzongkha in Bhutan is explained; they focus on description of the distribution and variation of the nasal and rhotic features, their sociolinguistic profile in the community, and issues of possible language change in progress.
Chapter One Page | 3
1.2 Motivation and Aim of the Study Such goals complement language planning and policy (LPP) objectives to standardise
Dzongkha, simplify it, make it user friendly and, moreover, to “make Dzongkha the main medium of communication for every Bhutanese in order to promote harmony, cohesion and stability in the country” (DDC 1989: 1). The above variation and change goals support efforts to document and describe the languages of Bhutan (e.g., Nyenkha project: Linguistic and ethnographic documentation of Western dialects of Nyenkha spoken in the Phobjikha valley in
Wangduephodrang, Bhutan [2019]), and contribute to our understanding of sociolinguistic practices in underdescribed speech communities in Asia. They also complement LPP objectives… Script reform, graphisation and modernisation are crucial LPP efforts in many
Asian countries but rarely studied in conjunction with variation and change. In the present case, lively debates on social media and in schools about whether to write final nasals and rhotics are evidence that LPP aims intertwine closely with variation and change.
1.2.1 The Context of Language Planning and Policy in Asia These are typical of contemporary LPP objectives in the region: for example, China has commissioned “China’s State Language Commission, an administrative department under the
Ministry of Education” since 2006, which compiles “an annual Green Paper on the so-called
‘language life’ in China”(Li 2015: 01-02). These papers are also published under the title
‘Language Situation in China’ since language is seen as the key to the ‘unity and identity’ of the
Chinese nation. Moreover, the Chinese people are tied to their linguistic history and ideology and the Ministry of Education encourages schools across the country to teach Chinese according to Chinese language policy and language planning. The Language Commission conducts
“standard test[s] for individuals who wish to hold public offices.” Hence, Chinese people are increasingly aware of their language policy, language planning and language rights (Li 2015: 01-
02).
Chapter One Page | 4
Likewise, as soon as India started exercising its self-government, “the question of which language to use for official purpose in the new nation sparked much discussion in light of the nation’s vast linguistic diversity” (Groff 2017: 10). Henceforth, Mahatma Gandi and a number of
Indian leaders decided to choose one of the nation’s languages as the official language, including
English to maintain efficiency in the government’s system. There are some requirements for an official language, as listed below (Groff 2017: 10):
1. It should be easy to learn for government officials,
2. It should be capable of serving as a medium of religious, economic and political
intercourse throughout India,
3. It should be the speech of the majority of the inhabitants of India,
4. It should be easy to learn for the whole of the country.
In this regard, “Hindustani, encompassing both Hindi and Urdu was promoted as the official language of India before the partition of India and Pakistan”. However, Hindi, written in the Devanagari script was declared as “the official language of the new nation, with English as an auxiliary official language” (Groff 2017: 10-12).
Chand (2013) also states that the Hindi was recognized as the official language of India,
“with 15 vernaculars designated as scheduled (officially sanctioned and funded) languages of government” (p. 858-59). Moreover, “Indian states were constitutionally authorized to establish their own state-based official language(s), resulting in an array of officially sanctioned languages and scripts” (Chand 2013: 859). Thus, Hindi, a standardized and Sanskritised register, is the lingua franca of Hindi land (Hindi belt), taught in schools, and used in courts and military offices.
As Shresta (2017) explains, Nepal is a multilingual country as “there are 123 languages and 125 castes and ethnic groups” according to the recent census (2011) in the country (p. 1).
Since language planning is inevitable for the Nepali government which is the arbiter of national
Chapter One Page | 5 identity, the first language planning and policy in Nepal was established in 1905. “Then, Nepali language was made as the language of law and government” with English language education in the government-run English medium school in a few places of Nepal (Shresta 2017: 03).
“However, Hindu Pathshalas and Baudha Gompas were using Sanskrit and Tibetan respectively as medium of instruction from the time immemorial in Nepal” (ibid. p. 05).
According to Chand (2013), “Nepal is a multilingual and pluralistic nation with 104-plus living languages with at least six scripts, and a distinct Nepali Sign Language” (p. 874). Nepali
(or Nepalese) also holds a crucial role in unifying the nation “through three recent socio-political periods and continuing today” (Chand 2013: 874). Likewise, the 1990 constitution acknowledged Nepali with Devanagari script as the official language and lingua franca in the country “while it permitted (yet did not fund) developing indigenous languages, cultures, and scripts and establish[ed] primary schools in various mother tongues” (ibid. p.875). Hence, Nepali is spoken throughout the nation, taught in schools, used in courts and military workplaces. It is also known as Khas language and belongs to Eastern Pahari, a sub-branch of the family of Indo-
Aryan languages, which has three branches: Jumli, Nepali and Palpa. Familiar stages in a model of language planning and policy, including several of the nations above, are outlined in such works as Haugen (1983), Spolsky (2004), Shohamy (2006), Ricento (2006), Cassels Johnson
(2013).
The purpose of this study is to describe language variation (and, where appropriate, signs of change) among Bhutanese people in spoken Dzongkha, and to obtain a fuller picture of the use, social distribution, and change in Dzongkha features among Eastern-Dzongkha speakers,
Western-Dzongkha speakers and Southern-Dzongkha speakers in the capital city.
The survey examines the realisation of nasality and rhoticity in reading and speaking across the residents of the capital city – typical topics for sociolinguistic variation research (e.g.
Barale 1982 on nasality in Mandarin; Abbas 2009 on nasality in Saraiki; Labov 1966 on
Chapter One Page | 6 rhoticity). Speakers come from different social categories, e.g. age, education and sex: for example, primary students (the young), primary school teachers (different ages and levels of seniority), primary students’ parents (different occupational classes and ages), all evenly divided between the sexes. Crucially, they also come from different ethnic and regional groups, typically originating in the western (Ngalong in Dzongkha), southern (Lhotsham in Dzongkha) and eastern (Sharchog in Dzongkha) regions. In some of these regions Dzongkha is typically not the first language of many adult speakers, though most people speak it as a second language.
Ngalong is the western and northern part of Bhutan, including Thimphu, and has been a
Dzongkha-speaking region for centuries. Lhotsham (which means “Southern” in Dzongkha) is the southern part of Bhutan, and a Nepali-speaking region.1 Sharchog (“Eastern” in Dzongkha) is the eastern part of Bhutan and linguistically diverse; however, Tshangla is a dominant language, and all speakers in my sample from Sharchog who do not speak Dzongkha natively are first- language Tshangla speakers (van Driem 1993, Dorji 1990, DDC 1999, Tshewang 2013). All speakers were born and reside in Bhutan and use Dzongkha as their official language, i.e. the language in which they interact with government and authorities.
As an important member of the Tibeto-Burman family, Dzongkha also provides an interesting subject for researchers. Measuring the attrition of final nasal consonants and post- vocalic rhoticity in spontaneous Dzongkha speech may reveal interesting characteristics of an understudied language, and in future might help clarify its relation to the languages of South
Asia in general. This research also aims to understand the nature of nasalization versus denasalization and rhotacization versus derhotacization processes, by employing the sociolinguistic research methods described in chapter three.
1 In this thesis I refer to it as Nepali; in Bhutan, the term “Lhotshampa” is used, and the Ethnologue language reference tool (Simons & Fennig 2018) uses the term “Lhotshammikha” for this variety.
Chapter One Page | 7
1.3 Significance of the Study At present, as mentioned in (1.1 above), debates and controversies over whether to pronounce the -n and -r endings are growing among teachers, parents and, of course, students at schools in general and within the department of Dzongkha curriculum, Ministry of Education, in particular. This study which analyses [de]nasalization and [de]rhotacization in Dzongkha speech will be germane to the sociolinguistic explanation of the nasalization and rhotacization process in the Dzongkha curriculum, and may serve as a resource towards the teaching and learning of
Dzongkha. Understanding the power of language ideology in society, this study may provide an evidence base of actual everyday speech to inform future policy decisions.
This study hopes to contribute to our understanding of (socio)linguistics by presenting an empirical analysis of the processes and relationships between language and society and related categories in Dzongkha. Especially, it may make a contribution towards the development of literature on Dzongkha and provide a vital piece of information for future scholars or linguists in this field.
It also aims to offer an up-to-date account of linguistic variation and change in Dzongkha at large. In addition, this study makes further documentation and analysis available on nasalization and rhotacization in Dzongkha for those who wish to do more study on them and make any official changes in the Dzongkha curriculum in the near future.
1.4 Structure of the Thesis This thesis is composed of six chapters. The first chapter presents the motivation and aims of the study in light of the linguistic situation in Bhutan. It provides a geographical, historical, and social overview of Dzongkha, including language policy in Bhutan. The linguistic history and classification of Dzongkha in Bhutan are described, and it concludes with research questions and main hypotheses.
Chapter One Page | 8
Chapter 2 presents an overview of the literature review, including the nature of the speech community, an overview and examples of language ideologies and behaviours, and studies of nasal and rhotic codas in Dzongkha. It also offers a brief description of the Dzongkha consonant and vowel inventories, and lexical tones in Dzongkha.
Chapter 3 discusses the research methods used in this study. It gives detailed information about the choice of research site, the structure of social relations within the community, choice of participants, their social characteristics and problems encountered during the research in the field. It also provides information about the stratification of the sample and its size, and how the researcher gained access to the speech community. Additionally, this chapter gives detailed information about what kind of recording instruments and analysis tools have been used, including research questionnaires and materials, sociolinguistic interviews, transcription protocols, coding techniques and the software used for data analysis.
Chapter 4 explores linguistic variation among nasal finals in Dzongkha, considering both internal linguistic constraints and speaker variables. It also gives more detailed information about the quantitative analysis of this linguistic variable along with the results of the patterns observed.
Chapter 5 explores linguistic variation for postvocalic (r) in Dzongkha, again investigating internal linguistic constraints as well as social factors. It also gives more detailed information about the quantitative analysis of this linguistic variable along with the results of the patterns observed.
Finally, chapter 6 reports a summary of the core findings, relating them to the research questions and hypotheses. Contributions to sociolinguistic research are discussed, and limitations of the study and suggestions for further research are given.
1.5 Research Questions and Hypothesises All Dzongkha speakers are well aware that there is variation and dialectal difference in the oral realisation of final nasals and rhotics. These variable features are frequently commented
Chapter One Page | 9 on in classrooms and in discussions of writing and speech. In addition, everyone who has studied
Dzongkha and/or Chöké is taught that final rhotics are to be pronounced in literary and high written forms. The fact that they are often not pronounced in the everyday speech of Thimphu today suggests strongly that change may be occurring. This wide popular awareness underlies the decision to study these features as possible changes in progress, employing the variationist paradigm. Thus, this sociolinguistic study addresses the important research questions:
1) What types of nasalization and rhoticity occur in Dzongkha?
2) Is there any deletion of nasal codas in Dzongkha by the residents of the capital city,
Thimphu? What linguistic factors constrain the deletion?
3) Is there any deletion of rhotic codas in Dzongkha by the residents of the capital city,
Thimphu? What linguistic factors constrain the deletion?
4) Have such changes already been completed, or are they still in progress?
5) Do eastern, western and southern Dzongkha-speakers differ in the realization of nasal
codas and rhoticity?
6) Do social factors such as sex, educational level, age and social class constrain the
realization of nasal codas and rhoticity?
Thus, the main hypotheses of the study are:
Ø H1: Realization of nasals in codas is hypothesized to show a systematic variation
constrained by system-internal linguistic factors. The list of factors is discussed in
Chapter 3, and their effects measured in Chapter 4.
Ø H2: Realization of rhoticity in codas is hypothesized to show a systematic variation
constrained by system-internal linguistic factors. The list of factors is discussed in
Chapter 3, and their effects measured in Chapter 5.
Ø H3: It is hypothesized that variable nasal and rhotic codas are changes in progress, thus
significant differences among age groups are expected. (However, the design of the study
Chapter One Page | 10
will not allow definitive results that might establish the nature of ongoing change in
detail.)
Ø H4: It is hypothesized that Dzongkha itself shows dialectal variation in the realisation of
nasal finals and rhoticity among groups of different geographical and ethnic origins
which involve local accents and localised variants.
Ø H5: It is hypothesized that Dzongkha shows sociolinguistic variation according to sex,
education, age and class, which are salient factors in acquiring linguistic competence.
Ø H6: It is hypothesized that language ideology and attitude, as well as stylistic context
(including interview, narrative and reading styles), influence patterns of variation. In
particular it is expected that reading styles will show conservative effects on variation of
nasals and rhotics.
The remainder of this chapter provides useful background and historical material concerning the language ecology of Bhutan and the history of Dzongkha.
1.6 Diglossic Situation Although Dzongkha is the official language of the Kingdom of Bhutan, Chöké or classical Tibetan has also been used as the religious language in Bhutan, particularly for
Buddhist literature over the last few centuries. Chöké refers to the language of texts written in a
Tibetan script which prevailed widely across the area of the Tibet Plateau and some areas of eastern Central Asia, including Nepal, Sikkim, Ladakh and Bhutan (Rinchen 1972, Hasrat 1980,
Nado 1986, Tshewang 1995). Chöké and Dzongkha are historically related and structurally similar in a number of ways, although Chöké is not intelligible to a Dzongkha speaker without study.
Chöké was first introduced by the great master Guru Padmasambhava2 in the 8thcentury and his followers. It is depicted in the Padmasambhava’s treasured biography “Zanglingma”
2 Literally this means “Lotus-Bon”; also known as Guru Rinpoche, he was a 7th-century Buddhist master.
Chapter One Page | 11 revealed by Nyangrel Nyima Özer (1136-1204). Both languages are used by the same language community, but today Chöké is used only in texts for liturgical purposes. It is used in certain situations, for instance, in formal Buddhist education for monks or lay practitioners, in liturgical performances, dialectical purposes (e.g. formal debates), or other specific settings, but it is not used for commonplace conversation by ordinary speakers in any everyday situation.
As in other countries, many speech communities in Bhutan also use two or more varieties under different circumstances; for example, “In Baghdad the Christian Arabs speak a "Christian
Arabic" dialect when talking among themselves but speak the general Baghdad dialect, "Muslim
Arabic," when talking in a mixed group” (Ferguson 1959: 325). In Bhutan, both Chöké and
Dzongkha are written languages: Chöké had been taught in schools until the 1970s, and
Dzongkha began to be taught in schools from 1971 based on the needs of its users, i.e. modernisation, and guided by the Royal edicts (Dorji 2009).
In addition, the diglossic situation in Bhutan is seen as a kind of bilingualism in society.
The two languages are closely related. Traditionally, Chöké has high prestige (referred to as
“H”) and Dzongkha has low prestige (referred to as “L”). They share a number of similar linguistic features and morphosyntactic processes at grammatical level, and Chöké is only partially intelligible to a person who is well versed in Dzongkha, unless they have also studied
Chöké. Most of the Chöké root letters remain unchanged in Dzongkha orthography, which requires replacing the subscribed letter in order to get the Dzongkha pronunciation and meaning, as Tshewang (1971) explains: “ར་འདོགས་ཅན་ནི་ཡ་འདོགས་ ར།” Ra dog cen ni ya dog jur /Rʌ-doɡ tʃʌn ni jʌ- doɡ ʒur/ which means, “Replace the /R/ subscribed letter with /Y/ subscribed letter”, so that e.g.